Br0kenTeleph0n3

Following the broadband money

How the UK’s broadband numbers stack up, or not

with 37 comments

BT published its numbers for 2011 this week. It revealed that its fibre “passes” 10 million homes. Of those, some 550,00 have subscribed to its Infinity VDSL fibre to the cabinet (FTTC) service.

BT says it added 589,000 retail broadband customers. It claims this is 54% of the net additions of 1,085,000, and took its retail broadband customer base to around 6.3m, up 10% in the year. Ofcom’s Communications Market Report, published in August 2011, said, “More than two-thirds (67%) of households have a fixed broadband connection and 17% have a mobile broadband (dongle) connection.”  That’s about 20 million homes, which puts BT’s share of the fixed broadband market at around 32%.

Despite doubling the network frequency in the local loop, which should have pushed its top download speeds from 40Mbps to 80Mbps, BT did not refer to it in its report. Nor does it respond substantively to questions about the  increase in speed actually experienced by customers.

The best indication of what customers might expect comes from Digital Region, the South Yorkshire wholesale broadband provider that is reportedly fighting for its financial existence due to earlier technology problems and a lack of customers. Digital Regional hired BT and Thales to provide its dedicated VDSL FTTC infrastructure. (This was an unhappy experience for reasons that we shan’t go into here.)

In early April, just after BT’s official speed upgrade, one of its resellers, the ISP Origin, provided its upload and download speeds to Thinkbroadband. The graph shows clearly how speed degrades with distance from the street cabinet.

The graph is not a smooth line because lots of things affect the user experience. They include distance to the cabinet, quality of the metal path (not every line is good quality copper), congestion on the link between the cabinet and the exchange, internal wiring in the customer premises, whether the modem or router is directly connected to the wall plug, whether the user is using Wi-Fi to connect, etc. The speed experienced by any individual is also increasingly affected by how many people in the premises are trying to connect at the same time, the so-called multiscreen syndrome.

The graph shows that BT may be underselling its VDSL technology by claiming a top download speed of “up to” 80Mbps. The Origin figures suggest that anyone living within 500m of their street cabinet should get an 80/20Mbps download/upload service, and anyone closer in should do better. That is if they, like Origin, have a dedicated rather than a shared service.

Points of presence of Digital Region’s dedicated VDSL FTTC service.

As Origin puts on its website, “BT’s cabinets are shared between all BT customers, Sky customers, PlusNet Customers and any other ISPs that piggyback on their fibre. So you’ve got loads of customers all wanting the same superfast speeds at the same time all using the same cabinet. This means that to make sure customers all get a half decent service, the ISPs have to limit and manage the speeds.

“We don’t like that. Our boxes are only used by Digital Region and we don’t fiddle with people’s connections, as everyone deserves the speeds they signed up for, surely?”

BTW, if you agree with that sentiment, you can join the 700-plus people who have singed Richard Brown’s petition for Ofcom to ban “up to” advertising here.

Advertisements

Written by Br0kenTeleph0n3

2012/05/13 at 15:16

37 Responses

Subscribe to comments with RSS.

  1. Report is to 31 March 2012, 40/80 went live in April.

    Has anyone come up with any proposals for pricing and marketing ADSL reflecting speed, throughput, usage etc?

    Somerset

    2012/05/13 at 15:39

    • You mean a flat rate per megabyte sent and received? Interesting idea. The mobile operator do that all the time. I suspect most fixed broadband consumers like fixed price contracts because it helps them budget. It would certainly get the bandwidth hogs such “online pirates” to pay their way, if telcos could find a definitive way to bill the right person.

      Ian Grant

      2012/05/13 at 17:32

      • “if telcos could find a definitive way to bill the right person.”

        Urm hardly rocket science, most providers can easily tell with fixed line broadband who is downloading and uploading in terms of amounts. How else do wll the 10GB, 20GB packages work.

        Andrew Ferguson

        2012/05/13 at 18:32

      • A court in the US recently held that an IP address does not guarantee identity and therefore liability to pay. The ACS-Law judge here held the same view. It’s one of the problems with the piracy provisions in the Digital Economy Act. Proving online identity has been a problem since before the New Yorker ran that dog cartoon, and it’s not solved yet.

        Ian Grant

        2012/05/13 at 22:19

      • The telco/ISP provides the IP address so it is clear who owns the line and using the data. You are confused with the far distant internet server which is not relevant here.

        Somerset

        2012/05/14 at 06:53

      • Who owns the line is not necessarily the same person who uses it, as anyone who kids have their own laptops or smartphones and a wireless router will attest. BTW, the courts in the UK and the US take this view, that an IP address is not proof of the user’s identity, no matter how much BPI, RIAA and MPAA would like it so.

        Ian Grant

        2012/05/21 at 23:53

      • If an ISP cannot analyse its radius logs and monitor traffic levels for a customer, then it does not deserve to be an ISP.

        Providers are able to successfully track usage against a subscribers line.

        What the IP address does not address who is actually using the data. So the copyright infringement cases are subtly different, as you need to identify the person at the keyboard to serve them with the notice.

        Andrew Ferguson

        2012/05/14 at 08:01

      • The person that ‘owns’ the line is the person that pays the rental and would pay for usage. Simple.

        Somerset

        2012/05/22 at 06:35

    • ‘It claims this is 54% of the net additions of 1,085,000’.

      Why the word ‘claims’, is there some doubt?

      Somerset

      2012/05/14 at 06:56

  2. I think there are enough real world results on VDSL speed so not sure what the point of this is. Crosstalk, Real copper distance, copper type, chipset, modulation, etc all have an impact.

    Many vendors have published lab speeds and expected speeds and actual speeds. As more people take the service better results will be available. but also as more people take the service crosstalk will increase.

    From my own house which is less than 500m from the cab I get around 65M/20M. On the upto point, does that mean that those that get a 80M service shouldn’t because I can’t? I don’t think so. Personally speaking I think a cooling off period should be allowed that if you don’t get the speed you want then you can cancel with no obligations, it will drive operators to drive vendors and installations to improve their quality rather than just crank the top speed down and give up.

    Far far too many things affect speed and not just the line length of your copper. Many people get high speeds from a sync point of view only to find out the Internet speeds are slowed down by traffic management or other type of technologies.

    oh and coming soon -> G.FAST upto 1G over copper!

    neilmcrae

    2012/05/13 at 15:56

    • Thanks for the comment Neil. I think your cooling off idea is great. I think one or two ISPs may be tacitly offering converted contracts if their customers show consistent data rates below what they’ve been promised. Not that they shout about it 😉
      The point is really to explore (for those less genned up than you) where we are headed with VDSL in terms of actual, not lab results. That’s why I’ve been asking people what the user’s experience is, and why I mentioned potential speed bottlenecks under user’s control, such as using a Wi-Fi router for access.

      Ian Grant

      2012/05/13 at 17:26

  3. just one thing I should have stressed is that crosstalk is liable to really impact the top line speeds above 80M very aggressively, so as more users are added and crosstalk increases this will have a really big impact with users potentially seeing speeds reducing!

    neilmcrae

    2012/05/13 at 16:14

    • What’s the solution to crosstalk? Better filters?

      Ian Grant

      2012/05/13 at 17:16

      • Filters cannot discriminate between good and bad VDSL2. Hence the magic of vectoring that should mitigate the crosstalk from other VDSL2 lines.

        Vectoring is on the verge of roll-out from labs, so we can expect trials in 2 to 3 years, and speeds in excess of the current estimates, i.e. a lot closer to the VDSL2 theoretical graphs.

        Sometimes I wonder if I am the only soul who reads more than the press releases when it comes to broadband

        Andrew Ferguson

        2012/05/13 at 17:41

      • You may be the only one here who does broadband full-time.

        Ian Grant

        2012/05/13 at 17:45

      • You don’t need to do broadband ‘full time’ to answer some of the questions and statements that appear on the subject of broadband. But I do suspect things like the ‘up to’ petition are there to cause annoyance for whatever reason. If they were not then there might be just a very slight hint at what a better pricing model would be. So far not one word apart from Ian’s reply above.

        So there is an issue with the wrong people being billed, please tell us more.

        Somerset

        2012/05/13 at 18:06

      • Vectoring, which makes a substantial difference but looks like an expensive solution.

        neilmcrae

        2012/05/13 at 19:15

  4. Also since when did the UK only have four providers? O2 anyone? Orange? and the list goes on.

    As for the Origin figures, I believe I did mention that as they are on a low uptake network, the effects of cross-talk are not taking effect. The other VDSL2 data I’ve done does allow for crosstalk at a fairly pessimistic level too.

    This will not be popular, but full fibre to the premises does not solve the up to either, because of contention. It very worrying when even IT firms forget this when campaigning against up to.

    Every week I come across people saying they are upgrading to faster service such as infinity so they can watch Netflix as their 5 Meg service cannot play it. If a 5 Meg service cannot, then most likely the problem is backhaul capacity.

    Andrew Ferguson

    2012/05/13 at 16:56

    • Never said they are the only ISPs, but they supply 18m out of the 20-odd million Ofcom counts as household receiving fixed broadband. O2 and Orange or even Everything Everywhere don’t seem to report their UK numbers, or at least I couldn’t find them easily. Perhaps you can provide a link.
      Re cross-talk and its effect on speed – well, what does BT say about that? At what point does one hit the law of diminishing return in terms of Mbps/£ invested? Surely that’s the clincher for FTTH?

      Ian Grant

      2012/05/13 at 17:15

      • BT say, presume you mean Openreach. Well they give an estimate for each postcode in the UK. Given enough reason to do so I could analyse the spread sheets, ut a weeks work for what benefit?

        http://www.thinkbroadband.com/news/5161-how-fast-will-fttc-run-at.html is about as good as you get in terms of easy summary.

        O2 Numbers
        http://www.ispreview.co.uk/index.php/2012/05/isp-o2-uk-and-be-see-slower-decline-to-617800-broadband-subscribers.html

        Orange Numbers (next quarter due any day)
        http://www.thinkbroadband.com/news/5061-orange-holds-its-broadband-customer-numbers-steady.html

        To exclude the other 7 to 9% of UK broadband users is doing a disservice to the many smaller providers, who provide the wide and varied market that is the UK broadband market.

        Andrew Ferguson

        2012/05/13 at 17:30

      • Thanks for the links Andrew, even if they do drive traffic to your site 😉
        Here’s what EE says in its actual press release: “We have also successfully migrated fixed broadband customers from a legacy network to BT’s network, which has lowered operating costs.”
        I don’t mean to diss smaller ISPs, but nearly all resell capacity from the four large ISPs and other carriers such as CWW, Geo, etc.
        Of course, O2 and BE and Zen regularly top user surveys of “the best broadband supplier”, and more strength to their elbows.
        Sadly, their ability to influence government, local councils’ or even their supplier’s decisions on new broadband infrastructure is, I suspect, negligible.

        Ian Grant

        2012/05/13 at 17:43

      • On the reselling capacity, the providers do have a different variety of ways of varying their QoS, e.g. traffic management, usage allowances, or unlimited. Just because someone uses BT Wholesale does not mean their service will perform the same as another provider on BT Wholesale. An ISP even has control to some extent over the DLM that governs the sync speed

        Andrew Ferguson

        2012/05/13 at 18:35

      • With g.fast coming which will give 1GB on copper we are along way off diminishing returns!

        neilmcrae

        2012/05/13 at 19:42

      • Has BT/Openreach said it will offer g.fast?

        Ian Grant

        2012/05/13 at 22:25

      • G.Fast Openreach – not yet.

        Too early in the fibre roll-outs to commit to something that may not be needed if takeup grows following the patterns it did with original ADSL services.

        Andrew Ferguson

        2012/05/14 at 08:02

    • BT investing £2.5bn if we assume all going on FTTC (Which it is not, but for a simple five minute shake down that is all I will do)

      £2.5bn covering 2/3rds, which is around 17 million homes, or £147 per premise and speed estimates from http://www.thinkbroadband.com/guide/fibre-broadband.html

      Then for the 75% likely to connect at 38Mbps or higher its £3.90 per Mbps.
      For the next 23% which we will worst case them at 15 Mbps irs £9.80

      The remaining 2%, can be ignored for such a rough and ready reckoner.

      Andrew Ferguson

      2012/05/13 at 17:37

      • That 2% includes people like the B4RN communities, which is why they’re JFDI themselves. On that score, any news on Defra’s rural community broadband fund? Anyone got any money via that source yet?

        Ian Grant

        2012/05/13 at 17:49

      • The Defra stuff just started last week, as in calls for registration of interest etc

        Andrew Ferguson

        2012/05/13 at 19:23

  5. […] See the original post here: How the UK's broadband numbers stack up, or not « Br0kenTeleph0n3 […]

  6. 1. For comparison purposes, it would be helpful for some figures for Virgin Media’s faster broadband services, although I appreciate they are still in the process of deploying more upgrades.

    2. Another important point for anybody striving to improve their connection is that they can be in for a very torrid time in obtaining satisfactory assistance from the infrastructure provider, particularly as the EU is not supposed to have access to the modem’s statistics.

    Walter

    2012/05/13 at 18:52

    • From VM’s 1Q12 report: “…we again saw around half of new subscribers taking superfast speeds of 30Mb or more. We added 146,700 such customers during the quarter, taking the total to nearly 850,000, which is more than triple the number of a year ago. This includes over 250,000 on our 50Mb, 60Mb or 100Mb tiers.”

      Ian Grant

      2012/05/13 at 22:23

      • Virgin now only sells a 30 Mbps product, though plenty of moans that people don’t get those speeds. Which brings into question how Ofcom monitors providers speeds.

        50 Meg has been around for three years. So worth viewing in that light.

        Andrew Ferguson

        2012/05/14 at 08:05

  7. Oh and we had a vectoring trial about 6 months ago 😉

    neilmcrae

    2012/05/13 at 19:17

    • How did it go? Can you give us details?

      Ian Grant

      2012/05/13 at 22:23

      • It worked as we expected it too! 🙂

        neilmcrae

        2012/05/22 at 06:44

  8. Looking back at the ‘Origin/TBB’ graph, is it actually saying that upload speeds at 1300m+ are pretty much zero?

    Supplementary: what is the ‘latest’ on the speed required to ‘qualify’ for an Infinity connection? I recall talk of dropping it to 5mb but I think it settled at 15?

    Ian – I believe roughly half the initial bids were ‘accepted’. Any involving fixed wireless were ‘deferred’ to the current round (which did open on 10/5) due to the on-going UK/EU ‘debate’ over wireless in terms of state aid.

    Mike Phillips

    2012/05/21 at 14:54

    • Mike, I understand there has been a massive fight between Europe and BDUK over what constitutes an acceptable speed for broadband. Aparently all local authority bids were based on BT’s “up to 24Mbs”. This was unacceptable to the commission, which is why it would not rubber-stamp the BDUK framework broadband procurement agreement.
      Europe insisted that the target is 30Mbps for all rather than 24Mbps, but it has conceded that it is too late to upgrade some early broadband projects that have already received funding, such as Cornwall.
      Whether the European Commission is happy with the UK’s complaisance with “up to” estimates rather than a hard 30Mbps is still unclear.

      Ian Grant

      2012/05/21 at 23:48


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: